In today’s competitive banking and insurance market, companies are constantly seeking innovative strategies to enhance clients’ satisfaction and loyalty. According to industry reports, acquiring a new customer can cost 5 times more than retaining an existing one.1 Leveraging AI offers a transformative approach to meet these goals with models predicting customer retention, and optimize upselling and cross-selling opportunities.
Machine learning models can analyze vast amount of data on customers tracking demographics, needs, interactions with the company’s people and systems, and understand/predict behavioral patterns that may result in churn. These insights empower insurers when dealing with the client or defining commercial strategies, but taking proactively the right action requires to understand how to use model’s results when talking to an individual or taking strategic decision.
So how to overcome these limitations?
In Intellico we do believe that Explainable AI can enhance decision-making. That’s why we embedded an Explainable AI engine in the solution Grapho4Upselling, for client retention also in the banking and insurance sector.
Explainable AI techniques return to the decision-maker the lists of key-drivers that are impacting the client’s behavior so that they can decide which promotion or pricing point apply and how to better lead the conversation with the client.
Let’s see how Grapho4Upselling works.
Grapho4Upselling: Explainable AI by Intellico for the banking and insurance sector
Grapho4Upselling is a solution aimed to get insights out of the relevant data and knowledge built all over time on the interactions with clients.
Grapho4Upselling has been configurated to support the agents in pre-assessing the likelihood of conversion during the interaction with the clients.
Grapho4Upselling offers 3 functionalities to the agents:
- Explorer: The agent visualizes the history of all the negotiations with the client, represented as a graph of interactions. The agent easily understands the number of appointments made in the past, their outcome and navigate the details regarding the type of needs discussed as well as the issued proposal.
- Explainer: The agent can access a dashboard highlighting the key drivers that, according to the AI model, better explains the behavioral pattern of the clients. Are they driven by specific needs or price ranges? How much is important to be proactive and periodically visit the client?
- Simulator: The agent simulates the impact of new conditions or testing different approaches in the interaction, based on the insights gathered with the Explorer and Explainer on past interactions.
What about the benefits of Grapho4Upselling?
The solution has been applied to model client’s conversion and customer retention.
Specifically, it aids in identifying optimal upselling opportunities by analyzing customer profiles and purchasing history, identifying additional service and add-ons that better meet clients’ needs. This is a forward-thinking strategy that do not simply focus on sales but strengthens the customer relationship.
Moreover, Grapho4Upselling can be expanded to suggest additional products for a cross-selling strategy according to the product portfolio of the specific clients. For instance, a customer with a home insurance might be interested in bundling it with welfare policy for a discount.
The agents that tested the solution experienced:
- Smooth Navigation across data: with Grapho4Upselling, the agents can navigate across the history of the client represented in a graph, drilling-down with a click
- Improved interactions with the client: Grapho4Upselling supports the users in preparing and leading the interactions with client. Moreover, it can help identifying the most promising clients to prioritize the agents’ agenda
Actionable insights: Grapho4Upselling suggest the agents the key drivers for the client so that they can better focus and target the commercial offering. - Actionable insights: Grapho4Upselling suggest the agents the key drivers for the client so that they can better focus and target the commercial offering.
What’s behind Grapho4Upselling?
Intellico’s approach to Explainable AI is based upon Graph Neural Networks (GNNs).
Graph Neural Networks are a class of neural networks designed to operate on graph-structured data. Unlike traditional neural networks, which typically process data represented in a grid or vector format, GNNs can effectively capture the complex relationships and dependencies present in data, once they are represented in a graph consisting of nodes (representing entities like clients, appointments, proposals,..) and edges (representing relationships between entities).
By employing these advanced techniques, Intellico identifies patterns and relationships within customer data, enabling insurers to provide precise and tailored upselling opportunities. This ensures transparency as the AI’s decisions and recommendations are explainable and interpretable. Grapho4Upselling utilizes this cutting-edge technology to automatically resolve and visually present model results in an intuitive way, making it easier to understand the factors influencing predictions, even when numerous variables are at play.
Conclusions
AI is revolutionizing the banking and insurance industry by enhancing strategies regarding customer retention, upselling, and cross-selling strategies. Nevertheless, AI insights may not be immediately turned into actionable measures.
Intellico’s solution Grapho4Upselling, instead, with its 3 functionalities—Explorer, Explainer, and Simulator—leverages Explainable AI to help agents understand the key drivers behind the individual’s behavior, suggesting actions to perform.
Do you want to learn more on Explainable AI? Visit Intellico’s blog on the importance of Explainable AI https://intellico.ai/blog/why-is-explainable-ai-important-for-companies/
References
[1] Forbes, “Why Customer Retention is King: The Evolution of Retention Marketing”, 2020. Available at: https://www.forbes.com/sites/jerryjao/2014/11/19/why-customer-retention-is-king-the-evolution-of-retention-marketing-part-1/#:~:text=Research%20over%20the%20past%20few,20%25%20of%20your%20existing%20customers.
